Given the nonexistent-to-moderate results for most education interventions, from charters to vouchers to merit pay to class size, I've occasionally wondered whether education policy and school quality really do achieve measurable results once demographic differences are sorted out. So this post by Kevin Carey, which breaks out city-level data from the National Assessment of Education Progress, is useful stuff. For instance, take math proficiency among poor 4th graders (defined as those eligible for the National School Lunch program):
New York City: 31%
Boston: 24%
Charlotte: 23%
Austin: 22%
Houston: 22%
San Diego: 22%
Los Angeles: 15%
Chicago: 12%
Cleveland: 10%
DC: 7%
As Carey says, low-income 4th graders in New York are more than 4 times as likely to be proficient as low income 4th graders in DC. And that's in two cities with similar percentages of students below the poverty line.
Now, what's hard about education data is that there really is a lot to disaggregate. Is New York a more economically and racially integrated city? My understanding, from Dana, is that its schools, at least, are. Are there cultural differences we're not accounting for? Hard to know. But in the aggregate, the data certainly suggests that New York is doing a far better job educating kids than, say, DC, and that they're seeing real achievement results as reward for their efforts.
